PPT Slide
- Set of samples St = {<l1, p1>, … <lN, pN>} described by position l and weight p
- Initialize sample set S0 according to prior knowledge
- For each motion D do:
- Sampling: Generate from each sample in St-1 a new sample according to motion model
- For each observation s do:
- Importance sampling: Re-weight each sample with the likelihood
- Resampling: Draw N samples from sample set St according to their likelihood
Monte Carlo Localization (SIR)
[Dellaert et al., 99],[Fox et al., 99]